With changes in lifestyle and dietary structure, the disease burden of atherosclerotic cardiovascular disease (ASCVD) in China has been increasing and shows a trend of younger onset. Dyslipidemia and inflammatory reactions play a key role in the disease progression. However, both lipid markers and inflammatory markers are multidimensional and interrelated, making it challenging to directly assess their relationship. Therefore, this study used canonical correlation analysis (CCA) to deconstruct the correlation between the two sets of variables, with the aim of providing new insights into the prevention and intervention of ASCVD.

To explore the canonical correlations between lipid ratios and inflammatory markers in individuals at risk of cardiovascular disease.

This study selected 7 518 individuals who underwent health check-ups at the Health Management Center, Tianjin Medical University General Hospital in 2020. A questionnaire was used to collect sociodemographic information, lifestyle, medical history, and family history of cardiovascular disease. Anthropometric measurements included height, weight, waist circumference (WC), and blood pressure, and BMI was calculated. Fasting venous blood was drawn for biochemical tests. Based on the China-PAR model for assessing the 10-year risk of ASCVD, the subjects were divided into three groups: low-risk (n=4 718), medium-risk (n=1 575), and high-risk (n=1 225). CCA was used to explore the correlation between lipid ratios[total cholesterol (TC) /high-density lipoprotein cholesterol (HDL-C), triglyceride (TG) /HDL-C, low-density lipoprotein cholesterol (LDL-C) /HDL-C] and inflammatory markers[neutrophil-to-lymphocyte ratio (NLR), neutrophil-to-monocyte ratio (NMR), lymphocyte-to-monocyte ratio (LMR), monocyte-to-HDL-C ratio (MHR), C-reactive protein (CRP) ].

A total of 7 518 subjects were included, with a median age of 49.00 (40.00, 57.00) years and a median 10-year ASCVD risk of 3.22% (1.02%, 7.51%) ; 4 133 (54.97%) were male and 3 385 (45.03%) were female. There were statistically significant differences in TG/HDL-C, TC/HDL-C, LDL-C/HDL-C, NLR, NMR, LMR, MHR, and CRP among the different 10-year ASCVD risk groups (P<0.05). CCA identified three pairs of canonical variables to replace the original variables to study the correlation between lipid ratios and inflammatory markers, all of which indicated a correlation between the two (r=0.569, 0.100, 0.049; P<0.05). In the first pair of standardized canonical variables, LDL-C/HDL-C was positively correlated with MHR within a certain range; in the second pair, TC/HDL-C was positively correlated with CRP. Stratified analysis showed that in the low-risk group, all three pairs of canonical variables indicated a correlation between lipid ratios and inflammatory markers (r=0.573, 0.110, 0.060; P<0.05). In the medium-risk group, the first two pairs of canonical variables indicated a correlation (r=0.552, 0.142; P<0.05). In the high-risk group, the first two pairs of canonical variables indicated a correlation (r=0.496, 0.135; P<0.05). In the low-risk group, in the first pair of standardized canonical variables, LDL-C/HDL-C was positively correlated with MHR within a certain range; in the second pair, TC/HDL-C was positively correlated with LMR within a certain range. In the medium-risk group, in the first pair of standardized canonical variables, LDL-C/HDL-C was positively correlated with MHR within a certain range; in the second pair, TC/HDL-C was positively correlated with LMR within a certain range. In the high-risk group, in the first pair of standardized canonical variables, TG/HDL-C was positively correlated with MHR within a certain range; in the second pair, TC/HDL-C was negatively correlated with NLR within a certain range.

Lipid ratios and inflammatory markers are canonically correlated, with a stronger correlation between LDL-C/HDL-C and MHR. LDL-C/HDL-C combines LDL-C and HDL-C to reflect their level changes and better indicates the lipid metabolism level in the body. MHR integrates the pro-inflammatory effect of monocytes and the anti-inflammatory effect of HDL-C, offering a greater advantage for predicting ASCVD. Future research needs more large-scale prospective studies to assess and validate their predictive performance, with the hope of providing new ideas for the assessment and intervention of ASCVD.

随着生活方式和饮食结构的改变，我国动脉粥样硬化性心血管疾病（ASCVD）的疾病负担日益加重，且呈现逐渐年轻化的趋势。脂质异常和炎症反应在疾病进程中起到了关键作用。然而血脂指标和炎症标志物均是多维度且相互关联的，直接评估二者之间的关系具有挑战性。因此，本研究采用典型相关性分析（CCA）解构两组变量之间的相关性，以期为ASCVD的防治干预提供新的思路。

探讨心血管疾病风险人群脂质比值与炎症指标之间的典型相关性。

选取2020年在天津医科大学总医院健康管理中心进行健康体检的7 518人为研究对象，采用问卷收集社会人口学信息、生活方式、既往疾病史以及心血管病家族史等信息，测量身高、体质量、腰围（WC）、血压，并计算BMI，采集清晨空腹静脉血进行生化检测。依据China-PAR模型评估ASCVD 10年发病风险，依据受检者10年ASCVD发病风险，分为低危组（n=4 718）、中危组（n=1 575）和高危组（n=1 225）。采用CCA解构脂质比值[总胆固醇（TC）/高密度脂蛋白胆固醇（HDL-C）、三酰甘油（TG）/HDL-C、低密度脂蛋白胆固醇（LDL-C）/HDL-C]与炎症指标[中性粒细胞与淋巴细胞比值（NLR）、中性粒细胞与单核细胞比值（NMR）、淋巴细胞与单核细胞比值（LMR）、单核细胞与HDL-C比值（MHR）、C反应蛋白（CRP）]的相关性。

共纳入受检者7 518人，中位年龄49.00（40.00，57.00）岁。中位10年ASCVD发病风险为3.22%（1.02%，7.51%），男4 133人（54.97%），女3 385人（45.03%），不同10年ASCVD发病风险组研究对象TG/HDL-C、TC/HDL-C、LDL-C/HDL-C、NLR、NMR、LMR、MHR、CRP比较，差异有统计学意义（P<0.05）。CCA共得出3对典型相关变量来代替原始变量研究脂质比值与炎症指标之间的相关性，结果显示，3对典型相关变量均认为脂质比值与炎症指标相关（r=0.569、0.100、0.049，P<0.05）；在第1对标准化典型变量中，在一定范围内，LDL-C/HDL-C与MHR呈正相关；在第2对标准化典型变量中，TC/HDL-C与CRP呈正相关。分层分析结果显示，在低危组中，3对典型相关变量均认为脂质比值与炎症指标正相关（r=0.573、0.110、0.060，P<0.05）；在中危组中，前2对典型相关变量认为脂质比值与炎症指标正相关（r=0.552、0.142，P<0.05）；在高危组中，前2对典型相关变量认为脂质比值与炎症指标正相关（r=0.496、0.135，P<0.05）。低危组：在第1对标准化典型变量中，在一定范围内，LDL-C/HDL-C与MHR呈正相关；在第2对标准化典型变量中，在一定范围内，TC/HDL-C与LMR呈正相关。中危组：在第1对标准化典型变量中，在一定范围内，LDL-C/HDL-C与MHR呈正相关；在第2对标准化典型变量中，在一定范围内，TC/HDL-C与LMR呈正相关。高危组：在第1对标准化典型变量中，在一定范围内，TG/HDL-C与MHR呈正相关；在第2对标准化典型变量中，在一定范围内，TC/HDL-C与NLR呈负相关。

脂质比值与炎症指标典型相关，其中LDL-C/HDL-C与MHR具有较强的正相关关系。LDL-C/HDL-C将LDL-C与HDL-C两者结合起来，同时反映其水平变化，能更好地反映体内血脂代谢水平。MHR整合了单核细胞的致炎和HDL-C的抗炎作用，对于AS的预测更具优势。在未来的研究中，需要更多大规模的前瞻性研究来评估验证其预测性能，以期为ASCVD评估干预提供新的思路。